Heller v. Doe 509 U.S. 312 (1993) Justice Kennedy delivered the opinion of the Court. At issue here are elements of Kentucky's statutory procedures, enacted in 1990, for the involuntary commitment of the mentally retarded. In many respects the procedures governing commitment of the mentally retarded and the mentally ill are parallel. The statutes recognize a large class of persons who can petition for an individual's involuntary commitment, whether on grounds of mental retardation or mental illness. Upon filing of the petition, the trial court must appoint counsel to represent the individual in question, unless he retains private counsel. The trial court also must examine the person who filed the petition and, if there is probable cause to believe that the individual who is the subject of the petition should be involuntarily committed, the court must order his examination by two qualified professionals. The subject of the proceeding has the right to retain a professional of his own choosing, who may "witness and participate in any examination" of him. In cases of commitment for mental retardation, a professional retained by the subject's "parent or guardian" also must be permitted to witness and participate in any examination. If both qualified professionals certify that the individual meets the criteria for involuntary commitment, the trial court must conduct a preliminary hearing. At the hearing, the court must receive as evidence the reports of these two professionals and any other professional retained under the statute. The individual whose commitment is sought may testify and may call and cross examine witnesses. In cases of mental retardation, at both the preliminary hearing and, if there is one, the final hearing, Kentucky law provides particular rights to guardians and immediate family members: "Guardians and immediate family members of the respondent shall be allowed to attend all hearings, conferences or similar proceedings; may be represented by private counsel, if desired; may participate in the hearings or conferences as if a party to the proceedings; may cross examine witnesses if desired; and shall have standing to appeal any adverse decision." § 202B.160(3) If the trial court determines that there is probable cause to believe that the subject should be involuntarily committed, it proceeds to a final hearing. At the final hearing, the State, through the county attorney for the county in which the person subject to the proceeding lives, prosecutes the petition, and counsel for the person defends against institutionalization. At this hearing, "[t]he manner of proceeding and the rules of evidence shall be the same as those in any criminal proceeding." As in the preliminary hearing, the subject of the proceedings may testify and call and cross examine witnesses. In proceedings for commitment based on mental retardation, the standard of proof is clear and convincing evidence, for mental illness, the standard is proof beyond a reasonable doubt. For commitment of the mentally retarded, four propositions must be proven by clear and convincing evidence: "that: (1) The person is a mentally retarded person; (2) The person presents a danger or a threat of danger to self, family, or others; (3) The least restrictive alternative mode of treatment presently available requires placement in [a residential treatment center]; and (4) Treatment that can reasonably benefit the person is available in [a residential treatment center]." The criteria for commitment of the mentally ill are in substance identical, requiring proof beyond a reasonable doubt that an individual "is a mentally ill person: (1) Who presents a danger or threat of danger to self, family or others as a result of the mental illness; (2) Who can reasonably benefit from treatment; and (3) For whom hospitalization is the least restrictive alternative mode of treatment presently available." Appeals from involuntary commitment proceedings are taken in the same manner as other appeals from the trial court. After enactment of the 1990 modifications, respondents moved for summary judgment in their pending lawsuit against petitioner. They argued, among other things, that the differences in treatment between the mentally retarded and the mentally ill--the different standards of proof and the right of immediate family members and guardians to participate as parties in commitment proceedings for the mentally retarded but not the mentally ill--violated the Equal Protection Clause's prohibition of distinctions that lack a rational basis, and that participation by family members and guardians violated the Due Process Clause. The District Court for the Western District of Kentucky accepted these arguments and granted summary judgment to respondents on these and other grounds not at issue here, 770 F. Supp. 354 (1991), and the Court of Appeals for the Sixth Circuit affirmed, Doe v. Cowherd, 965 F. 2d 109 (1992). We granted Kentucky's petition for certiorari, 506 U. S. ___ (1992), and now reverse. Respondents contend that, in evaluating the constitutionality of the distinctions drawn by Kentucky's statutes, we should apply not rational basis review, but some form of heightened scrutiny. This claim is not properly presented. Respondents argued before the District Court and the Court of Appeals only that Kentucky's statutory scheme was subject to rational basis review, and the courts below ruled on that ground. Indeed, respondents have conceded that they pressed their heightened scrutiny argument for the first time in their merits brief in this Court. Id., at 23 ("[R]espondents did not argue this particular issue below . . ."). Even if respondents were correct that heightened scrutiny applies, it would be inappropriate for us to apply that standard here. Both parties have been litigating this case for years on the theory of rational basis review, which, as noted below, does not require the State to place any evidence in the record, let alone the extensive evidentiary showing that would be required for these statutes to survive heightened scrutiny. It would be imprudent and unfair to inject a new standard at this stage in the litigation. See Tennessee v. Dunlap, 426 U.S. 312, 316, n. 3 (1976); Ernst & Ernst v. Hochfelder, 425 U.S. 185, 215 (1976). We therefore decide this case as it has been presented to the courts whose judgments are being reviewed. We many times have said, and but weeks ago repeated, that rational basis review in equal protection analysis "is not a license for courts to judge the wisdom, fairness, or logic of legislative choices." FCC v. Beach Communication, Inc., 508 U. S. ___, ___ (1993) (slip op., at 5-6). See also, e. g., Dandridge v. Williams, 397 U.S. 471, 486 (1970). Nor does it authorize "the judiciary [to] sit as a superlegislature to judge the wisdom or desirability of legislative policy determinations made in areas that neither affect fundamental rights nor proceed along suspect lines." New Orleans v. Dukes, 427 U.S. 297, 303 (1976) (per curiam). For these reasons, a classification neither involving fundamental rights nor proceeding along suspect lines is accorded a strong presumption of validity. Such a classification cannot run afoul of the Equal Protection Clause if there is a rational relationship between the disparity of treatment and some legitimate governmental purpose. Further, a legislature that creates these categories need not "actually articulate at any time the purpose or rationale supporting its classification." Instead, a classification "must be upheld against equal protection challenge if there is any reasonably conceivable state of facts that could provide a rational basis for the classification." A State, moreover, has no obligation to produce evidence to sustain the rationality of a statutory classification. "[A] legislative choice is not subject to courtroom fact finding and may be based on rational speculation unsupported by evidence or empirical data." A statute is presumed constitutional, and "[t]he burden is on the one attacking the legislative arrangement to negative every conceivable basis which might support it," Lehnhausen v. Lake Shore Auto Parts Co., 410 U.S. 356, 364 (1973) (internal quotation marks omitted), whether or not the basis has a foundation in the record. Finally, courts are compelled under rational basis review to accept a legislature's generalizations even when there is an imperfect fit between means and ends. A classification does not fail rational basis review because it " `is not made with mathematical nicety or because in practice it results in some inequality.' " Dandridge v. Williams, supra, at 485, quoting Lindsley v. Natural Carbonic Gas Co., 220 U.S. 61, 78 (1911)…We have applied rational basis review in previous cases involving the mentally retarded and the mentally ill. See Cleburne v. Cleburne Living Center, Inc., 473 U.S. 432 (1985)…In neither case did we purport to apply a different standard of rational basis review from that just described. True, even the standard of rationality as we so often have defined it must find some footing in the realities of the subject addressed by the legislation. That requirement is satisfied here. Kentucky has proffered more than adequate justifications for the differences in treatment between the mentally retarded and the mentally ill. Kentucky argues that a lower standard of proof in commitments for mental retardation follows from the fact that mental retardation is easier to diagnose than is mental illness. That general proposition should cause little surprise, for mental retardation is a developmental disability that becomes apparent before adulthood. See American Psychiatric Assn., Diagnostic and Statistical Manual of Mental Disorders 29 (3d rev. ed. 1987) (hereinafter Manual of Mental Disorders); American Assn. on Mental Retardation, Mental Retardation: Definition, Classification, and Systems of Support 5, 16-18 (9th ed. 1992) (hereinafter Mental Retardation); S. Brakel, J. Parry, & B. Weiner, The Mentally Disabled and the Law 16-17, 37 (3d ed. 1985). By the time the person reaches 18 years of age the documentation and other evidence of the condition have been accumulated for years. Mental illness, on the other hand, may be sudden and may not occur, or at least manifest itself, until adulthood. See, e. g., Manual of Mental Disorders 190 (onset of schizophrenia may occur any time during adulthood); id., at 220, 229 (onset of depression usually is during adulthood). Furthermore, as we recognized in an earlier case, diagnosis of mental illness is difficult. See Addington v. Texas, 441 U.S. 418, 430 (1979). See also Mentally Disabled 18. Kentucky's basic premise that mental retardation is easier to diagnosis than is mental illness has a sufficient basis in fact. See, e. g., id., at 16; Ellis & Luckasson, Mentally Retarded Criminal Defendants, 53 Geo. Wash. L. Rev. 414, 438-439 (1985). This difference between the two conditions justifies Kentucky's decision to assign a lower standard of proof in commitment proceedings involving the mentally retarded. In assigning the burden of proof, Kentucky was determining the "risk of error" faced by the subject of the proceedings. Addington v. Texas, supra, at 423. If diagnosis is more difficult in cases of mental illness than in instances of mental retardation, a higher burden of proof for the former tends to equalize the risks of an erroneous determination that the subject of a commitment proceeding has the condition in question… There is, moreover, a "reasonably conceivable state of facts," Beach Communications, 508 U. S., at ___ (slip op., at 6), from which Kentucky could conclude that the second prerequisite to commitment--that "[t]he person presents a danger or a threat of danger to self, family, or others," Ky. Rev. Stat. Ann. § 202B.040 (Michie 1991)--is established more easily, as a general rule, in the case of the mentally retarded. Previous instances of violent behavior are an important indicator of future violent tendencies. See, e. g., J. Monahan, The Clinical Prediction of Violent Behavior 71-72 (1981) (hereinafter Monahan); Kozol, Boucher, & Garofalo, The Diagnosis and Treatment of Dangerousness, 18 Crime & Delinquency 371, 384 (1972). Mental retardation is a permanent, relatively static condition, see Mentally Disabled 37, so a determination of dangerousness may be made with some accuracy based on previous behavior. We deal here with adults only, so almost by definition in the case of the retarded there is an 18 year record upon which to rely. This is not so with the mentally ill. Manifestations of mental illness may be sudden, and past behavior may not be an adequate predictor of future actions. Prediction of future behavior is complicated as well by the difficulties inherent in diagnosis of mental illness. Developments in the Law--Civil Commitment of the Mentally Ill, 87 Harv. L. Rev. 1190, 1242-1243 (1974). It is thus no surprise that psychiatric predictions of future violent behavior by the mentally ill are inaccurate. See, e. g., Steadman, Employing Psychiatric Predictions of Dangerous Behavior: Policy vs. Fact, in Dangerous Behavior: A Problem in Lawand Mental Health 123, 125-128 (C. Frederick ed. 1978); Monahan 47-49. For these reasons, it would have been plausible for Kentucky to conclude that the dangerousness determination was more accurate as to the mentally retarded than the mentally ill. There is a further, more far reaching rationale justifying the different burdens of proof: The prevailing methods of treatment for the mentally retarded, as a general rule, are much less invasive than are those given the mentally ill. The mentally ill are subjected to medical and psychiatric treatment which may involve intrusive inquiries into the patient's innermost thoughts, see Meissner & Nicholi, The Psychotherapies: Individual, Family, and Group, in The Harvard Guide to Modern Psychiatry 357-385 (A. Nicholi ed. 1978) (hereinafter Harvard Guide), and use of psychotropic drugs, see Baldessarini, Chemotherapy, in Harvard Guide 387-431; Berger, Medical Treatment of Mental Illness, 200 Science 974 (1978); Mentally Disabled 327-330; Brief for American Psychological Association as Amicus Curiae in Washington v. Harper, O. T. 1988, No.88-599, pp. 10-11. By contrast, the mentally retarded in general are not subjected to these medical treatments. Rather, " `because mental retardation is . . . a learning disability and training impairment rather than an illness,' " Youngblood v. Romeo, 457 U.S. 307, 309, n. 1 (1982), quoting Brief for American Psychiatric Association as Amicus Curiae in Youngblood v. Romeo, O.T. 1981, No. 80-1429, p. 4, n. 1, the mentally retarded are provided "habilitation," which consists of education and training aimed at improving self care and self sufficiency skills. See Youngblood, supra, at 309, n. 1; M. Rosen, G. Clark, & M. Kivitz, Habilitation of the Handicapped 47-59 (1977); Mentally Disabled 332. *** Kentucky's burden of proof scheme, then, can be explained by differences in the ease of diagnosis and the accuracy of the prediction of future dangerousness and by the nature of the treatment received after commitment. Each of these rationales, standing on its own, would suffice to establish a rational basis for the distinction in question. There is a rational basis also for the other distinction challenged by respondents: that Kentucky allows close relatives and guardians to participate as parties in proceedings to commit the mentally retarded but not the mentally ill. As we have noted, by definition, mental retardation has its onset during aperson's developmental period. Mental retardation, furthermore, results in "deficits or impairments in adaptive functioning," that is to say, "the person's effectiveness in areas such as social skills, communication, and daily living skills, and how well the person meets the standards of personal independence and social responsibility expected of his or her age by his or her cultural group." Manual of Mental Disorders 28-29. See also Mental Retardation 5-6, 1516, 38-41. Based on these facts, Kentucky may have concluded that close relatives and guardians, both of whom likely have intimate knowledge of a mentally retarded person's abilities and experiences, have valuable insights which should be considered during the involuntary commitment process. Mental illness, by contrast, may arise or manifest itself with suddenness only after minority, when the afflicted person's immediate family members have no knowledge of the medical condition and have long ceased to provide care and support. Further, determining the proper course of treatment may be far less dependent upon observations made in a household setting…To be sure, Kentucky could have provided relatives and guardians of the mentally retarded some participation in commitment proceedings by methods short of providing them status as parties. That, however, isirrelevant in rational basis review. We do not require Kentucky to have chosen the least restrictive means of achieving its legislative end. San Antonio Independent School Dist. v. Rodriguez, 411 U.S. 1, 51 (1973)… We turn now to respondents' claim that one aspect of the involuntary commitment procedures violates procedural due process. We note at the outset that respondents challenge as violative of due process only those provisions of Kentucky's comprehensive involuntary commitment procedures that allow participation in the proceedings by guardians and immediate family members…Respondents claim that by allowing the participation of persons whose interests may be adverse to those of the individual facing possible involuntary commitment, the statute "skews the balance" against the retarded individual and therefore imposes a burden on him. Brief for Respondents 32-36. Both courts below apparently accepted this argument, almost without explanation. See 965 F. 2d, at 113; 770 F. Supp., at 358. In our view, the claim is without merit. We evaluate the sufficiency of this procedural rule under Mathews v. Eldridge, 424 U.S. 319 (1976). There we held that determining the dictates of due process requires consideration of three factors: "First, the private interest that will be affected by the official action; second, the risk of an erroneous deprivation of such interest through the procedures used, and the probable value, if any, of additional or substitute procedural safeguards; and finally, the Government's interest, including the function involved and the fiscal and administrative burdens that the additional or substitute procedural requirement would entail." Id., at 335. We think that application of the Mathews v. Eldridge factors compels the conclusion that participation as parties by close relatives and legal guardians is not a deprivation of due process. Even if parents, close family members, or legal guardians can be said in certain instances to have interests "adverse to [those of] the person facing commitment," 965 F. 2d, at 113, we simply do not understand how their participation as formal parties in the commitment proceedings increases "the risk of an erroneous deprivation," 424 U. S., at 335, of respondents' liberty interest. Rather, for the reasons explained above, these parties often will have valuable information that, if placed before the court, will increase the accuracy of the commitment decision…Guardians have a legal obligation to further the interests of their wards, and parents and other close relatives of a mentally retarded person, after living with and caring for the individual for 18 years or more, have an interest in his welfare that the State may acknowledge. For example, parents who for 18 years or longer have cared for a retarded child can face changed circumstances resulting from their own advancing age, when the physical, emotional, and financial costs of caring for the adult child may become too burdensome for the child's best interests to be served by care in their home… To be sure, if the additional parties involved in the proceedings favor commitment, their participation may increase the chances that the result of the proceeding will be a decision to commit. That fact, however, is beside the point. "The Due Process Clause does not . . . require a State to adopt one procedure over another on the basis that it may produce results more favorable to" the party challenging the existing procedures. Medina v. California, 505 U. S. ___, ___ (1992) (slip op., at 13). "The function of legal process, as that concept is embodied in the Constitution, and in the realm of factfinding, is to minimize the risk of erroneous decisions. Because of the broad spectrum of concerns to which the term must apply, flexibility is necessary to gear the process to the particular need; the quantum and quality of the process due in a particular situation depend upon the need to serve the purpose of minimizing the risk of error." Greenholtz v. Inmates of Nebraska Penal and Correctional Complex, 442 U.S. 1, 13 (1979). *** Because allowing guardians and immediate family members to participate as parties in commitment proceedings increases the accuracy of those proceedings and implements the State's interest in providing family members a voice in the proceedings, without undermining those interests of the individual protected by the Due Process Clause, these Kentucky statutes do not run afoul of due process… In sum, there are plausible rationales for each of the statutory distinctions challenged by respondents in this case… The judgment of the Court of Appeals for the Sixth Circuit is Reversed. JUSTICE O'CONNOR, concurring in the judgment in part and dissenting in part. I agree with JUSTICE SOUTER that Kentucky's differential standard of proof for committing the mentally ill and the mentally retarded is irrational and therefore join Part II of his opinion. I conclude, however, that there is a rational basis for permitting close relatives and guardians to participate as parties in proceedings to commit the mentally retarded but not the mentally ill. As the Court points out, there are sufficiently plausible and legitimate reasons for the legislative determination in this area. I also agree with the Court that allowing guardians and immediate family members to participate as parties in commitment proceedings does not violate procedural due process. Like my colleagues, I would not reach the question whether heightened equal protection scrutiny should be applied to the Kentucky scheme. JUSTICE BLACKMUN, dissenting. I join JUSTICE SOUTER'S dissenting opinion, for I agree with him that this statute is not even rational. I write separately only to note my continuing adherence to the view that laws that discriminate against individuals with mental retardation, Cleburne v. Cleburne Living Center, Inc., 473 U. S. 432, 455 (1985) (opinion of Marshall, J., joined by Brennan and BLACKMUN, JJ.), or infringe upon fundamental rights, Foucha v. Louisiana, 504 U. S. 71, 84-86 (1992) (plurality opinion of WHITE, J., joined by BLACKMUN, STEVENS, and SOUTER, JJ.), are subject to heightened review. JUSTICE SOUTER, with whom JUSTICE BLACKMUN and JUSTICE STEVENS join, and with whom JUSTICE O'CONNOR joins [in part], dissenting. To begin with, the Court declines to address Doe's argument that we should employ strict or heightened scrutiny in assessing the disparity of treatment challenged here. [n.1] While I may disagree with the Court's basis forits conclusion that this argument is not "properly presented," ante, at 5, I too would decline to address the contention that strict or heightened scrutiny applies. I conclude that the distinctions wrought by the Kentucky scheme cannot survive even that rational basis scrutiny, requiring a rational relationship between the disparity of treatment and some legitimate governmental purpose, which we have previously applied to a classification on the basis of mental disability, see Cleburne v. Cleburne Living Center, Inc., 473 U.S. 432, 446447 (1985), and therefore I need not reach the question of whether scrutiny more searching than Cleburne's should be applied. Cleburne was the most recent instance in which we addressed a classification on the basis of mental disability, as we did by enquring into record support for the State's proffered justifications, and examining the distinction in treatment in light of the purposes put forward to support it. While the Court cites Cleburne once, and does not purport to overrule it, neither does the Court apply it, and at the end of the day Cleburne's status is left uncertain. I would follow Cleburne here. Obviously there are differences between mental retardation and mental illness. They are distinct conditions, they have different manifestations, they require different forms of care or treatment, and the course of each differs. It is without doubt permissible for the State to treat those who are mentally retarded differently in some respects from those who are mentally ill. The question here, however, is whether some difference between the two conditions rationally can justify the particular disparate treatment accorded under this Kentucky statute. The first distinction wrought by the statute is the imposition of a lesser standard of proof for involuntary institutionalization where the alleged basis of a need for confinement is mental retardation rather than mental illness…At issue in this case is only the application of this provision to adults who have not been shown to be mentally retarded, but who are simply alleged to be. The subject of such a proceeding retains as full an interest in liberty as anyone else. The State of Kentucky has deemed this liberty interest so precious that, before one may be institutionalized on the basis of mental illness, the statutory prerequisites must be shown "beyond a reasonable doubt." § 202A.076(2). However, when the allegation against the individual is one of mental retardation, he is deprived of the protection of that high burden of proof. The first question here, then, is whether, in light of the State's decision to provide that high burden of proof in involuntary commitment proceedings where illness is alleged, there is something about mental retardation that can rationally justify provision of less protection. In upholding this disparate treatment, the Court relies first on the State's assertion that mental retardation is easier to diagnose than mental illness. It concludes that the discrimination in burdens of proof is rational because the lessened " `risk of error' " resulting from the higher burden of proof…The Court reaches essentially the same conclusion with respect to the second prerequisite, that the individual present a danger or threat of danger to himself or others… In concluding, however, that the demands of minimal rationality are satisfied if burdens of proof rise simply with difficulties of proof, the Court misunderstands the principal object in setting burdens. It is no coincidence that difficult issues in civil cases are not subject to proof beyond a reasonable doubt and that even the most garden variety elements in criminal cases are not to be satisfied by a preponderance of evidence. The reason for this is that burdens of proof are assigned and risks of error are allocated not to reflect the mere difficulty of avoiding error, but the importance of avoiding it as judged after a thorough consideration of those respective interests of the parties that will be affected by the allocation. See Addington v. Texas, 441 U. S. 418 (1979), at 425 In a civil commitment proceeding, on the State's side of the balance, are the interests of protecting society from those posing dangers and protecting the ill or helpless individual from his own incapacities. On the other side, it is clear that "[i]n cases involving individual rights, whether criminal or civil, `[t]he standard of proof [at a minimum] reflects the value society places on individual liberty,'" id., at 425 (brackets in original and citation omitted), which encompasses both freedom from restraint and freedom from the stigma that restraint and its justifications impose on an institutionalized person. Id. at 425-426. …Both the ill and the retarded may be dangerous, each may require care, and the State's interest is seemingly of equal strength in each category of cases. No one has or would argue that the value of liberty varies somehow depending on whether one is alleged to be ill or retarded, and a mentally retarded person has as much to lose by civil commitment to an institution as a mentally ill counterpart, including loss of liberty to "choos[e] his own friends and companions, selec[t] daily activities, decid[e] what to eat, and retai[n] a level of personal privacy," among other things. Brief for American Association on Mental Retardation (AAMR), et al. as Amici Curiae 12 (AAMR Br.)…Even assuming, then, that the assertion of different degrees of difficulty of proof both of mental illness and mental retardation and of the dangerousness inherent in each condition is true (an assertion for which there is no support in the record), it lends not a shred of rational support to the decision to discriminate against the retarded in allocating the risk of erroneous curtailment of liberty. *** The Court also rests its conclusion on the view that "it would have been plausible for the Kentucky legislature to believe that most mentally retarded individuals who are committed receive treatment which is . . . less invasive tha[n] that to which the mentally ill are subjected." Ante, at 12. Nothing cited by the Court, however, demonstrates that such a belief would have been plausible for the Kentucky legislature, nor does the Court's discussion render it plausible now…One recent examination of institutions for the mentally retarded in Kentucky's neighboring State of Missouri, for example, found that 76% of the institutionalized retarded receive some type of psychoactive drug and that fully 54% receive psychotropic drugs. See Intagliata & Rinck, Psychoactive Drug Use in Public and Community Residential Facilities for Mentally Retarded Persons, 21 Psychopharmacology Bull. 268, 272-273 (1985)… The Court also suggests that medical treatment for the mentally retarded is less invasive than in the case of the mentally ill because the mentally ill are subjected to psychiatric treatment that may involve intrusive enquiries into the patient's innermost thoughts. See ante, at 11. Again, I do not disagree that the mentally ill are often subject to intrusive psychiatric therapy. But the mentally retarded too are subject to intrusive therapy, as the available material on the medical treatment of the mentally retarded demonstrates… The same sorts of published authorities on which the Court relies, in sum, refute the contention that "[t]he prevailing methods of treatment for the mentally retarded, as a general rule, are much less invasive than are thosegiven the mentally ill." Ante, at 11. The available literature indicates that psychotropic drugs and invasive therapy are routinely administered to the retarded as well as the mentally ill, and there are no apparent differences of therapeutic regimes that would plausibly explain less rigorous commitment standards for those alleged to be mentally retarded than for those alleged to be mentally ill. With respect to the involvement of family members and guardians in the commitment proceeding, the Court holds it to be justified by the fact that mental retardation "has its onset during a person's developmental period," while mental illness "may arise or manifest itself with suddenness only after minority." Ante, at 15-16. The Court suggests that a mentally ill person's parents may have "ceased to provide care and support" for him well before the onset of illness, whereas parents are more likely to have retained connection with a retarded son or daughter, whose "proper course of treatment" may depend on matters related to "observations made in a household setting." Ante, at 16. In any event, although these differences might justify a scheme in which immediate relatives and guardians were automatically called as witnesses in cases seeking institutionalization on the basis of mental retardation, they are completely unrelated to those aspects of the statute to which Doe objects: permitting these immediate relatives and guardians to be involved "as parties" so as to give them, among other things, the right to appeal as "adverse" a decision not to institutionalize the individual who is subject to the proceedings. Where the third party supports commitment, someone who is alleged to be retarded is faced not only with a second advocate for institutionalization, but with a second prosecutor with the capacity to call and cross examine witnesses, to obtain expert testimony and to raise an appeal that might not otherwise be taken, whereas a person said to require commitment on the basis of mental illness is not… The Court simply points to no characteristic of mental retardation that could rationally justify imposing this burden of a second prosecutor on those alleged to be mentally retarded where the State has decided not to impose it upon those alleged to be mentally ill. Even if we assumed a generally more regular connection between the relatives and guardians of those alleged to be retarded than those said to be mentally ill, it would not explain why the former should be subject to a second prosecutor when the latter are not. *** Without plausible justification, Kentucky is being allowed to draw a distinction that is difficult to see as resting on anything other than the stereotypical assumption that the retarded are "perpetual children," an assumption that has historically been taken to justify the disrespect and "grotesque mistreatment" to which the retarded have been subjected. See Cleburne, 473 U. S., at 454 (Stevens, J., concurring)…When the State of Kentucky sets up its respective schemes for institutionalization on the basis of mental illness and mental retardation, it too is obliged to reject that analogy, and to rest any difference in standards for involuntary commitment as between the ill and the retarded on some plausible reason. In the absence of any rational justification for the disparate treatment here either with respect to the burdens of proof or the participation of third parties in institutionalization proceedings, I would affirm the judgment of the Court of Appeals. Because of my conclusion, that the statute violates equal protection, I do not reach the question of its validity under the Due Process Clause.